1. Field
The present invention relates to a mobile robot, and more particularly to a method, medium, and apparatus for performing path planning of a mobile robot, which can make the mobile robot moving through a narrow area by using a coarse map and a fine map.
2. Description of the Related Art
Generally, a robot developed for use in the field of industry has been used as a part of factory automation or has been used to perform tasks on behalf of a human being under the condition that cannot be overcome by the human being. An example of the robot for home use includes a cleaning robot serving as a leading part that extends heavy industry based robot engineering limited to a robot for industry use to light industry based robot engineering. The cleaning robot includes a driving unit for movement, a cleaning unit for cleaning, and a position measurement unit for measuring the position of the robot itself or the position of a user remote controller.
In a mobile robot such as a cleaning robot, it is the most basic and important function to detect an accurate position of the robot itself. In calculating an absolute position of a mobile robot, a method using beacons which adopts an ultrasonic sensor and is mounted in the home and a method using a GPS (Global Positioning System) for indoor use have been known. In determining a relative position, a method of obtaining the position of a mobile robot by obtaining a rotating speed and a straight traveling speed of the mobile robot and integrating the obtained speeds, a method of obtaining the position of a mobile robot by twice integrating acceleration values obtained from an acceleration sensor, and a method of obtaining the direction of a mobile robot by integrating a rotating speed that is an output value of a gyroscope have been known.
In order for a mobile robot to perform a specified work such as cleaning in a robot movement space, it is required to prepare a map, and for this, separate algorithms for exploring an unknown area are required in addition to a SLAM (Simultaneous Localization And Mapping). Such algorithms correspond to a simple method of locating the whole configuration of the robot movement space through a wall-following and a complicated method of performing path planning by using an active SLAM. Also, even after the map is prepared through the above-described methods, a separate coverage path planning for the mobile robot to thoroughly cover the whole area included in the map is additionally required.
In order for the mobile robot to cover a wide area, a cell decomposition, which divides the whole area into a plurality of cells, makes the robot perform a cleaning of each cell, and then moves the robot to another cell to complete the cleaning of the whole area, has been widely used.
The mobile robot prepares two-dimensional (2D) map information of an environment by using an infrared ray (IR) sensor, an ultrasonic sensor, a laser sensor, and so forth. The map as prepared above is called a grid map. Since this grid map is obtained from sensor data, it has great uncertainty and error.
The conventional mobile robot, however, has the following problems.
In the case of a fine map, which divides the area by using precise grid map that has cells having a size of about 1˜2 cm and performs the path planning, in the cell decomposition, path planning for cleaning every nook and corner can be performed. However, in a state that the grid map for the whole area is not completed, the division of the area becomes difficult to cause an inefficient path planning to be performed. In addition, if a grid map having a high resolution is used, large amounts of memory capacity and computations are required, and thus it is impossible to perform the path planning in real time in an embedded system.
In order to solve the above-described problems, a method of performing path planning in a short time by using a coarse map composed of square cells having the same size of several tens of centimeters may be used. In this method, the moving path of the robot is controlled using a coarse map in the range of 20˜25 cm that corresponds to the size of a suction port of the mobile robot, and thus a real-time map preparation becomes possible. According to this method, however, since the unit of movement is in the range of 20˜25 cm and the coarse map is updated with such a degree of resolution, non-movement of the robot frequently occurs even in an area where the robot can move due to the actual size of the mobile robot, and thus the movement of the mobile robot in every nook and corner becomes impossible.